- Measure application latency, throughput and availability requirements before choosing an access circuit or dark fibre.
- True resilience depends on verified route, duct, entry-point, power and equipment diversity, not simply buying two circuits.
- Choose OM4, OM5 or OS2 from the Ethernet application, transceiver, distance and loss budget; there is no universal 300 m changeover rule.
The UK data centre boom: what it means for commercial buildings
UK demand for data-centre, cloud and edge capacity continues to increase, with significant activity around London and other regional clusters. For an individual building, the relevant issue is not the headline market size but the available carrier routes, service locations, diversity and measured performance to the chosen provider.
For commercial buildings, this matters because it changes the connectivity calculus. The question for an IT Director or Facilities Director is no longer just “how do we connect devices within the building”, it is “how does the building connect to the data centre infrastructure that hosts our AI compute, cloud services and business-critical applications?” The answer to that question starts with the last-mile connection between the building and the carrier network or dark fibre infrastructure that reaches the data centre.
What the last-mile connection looks like
The last-mile connection is the physical link between a commercial building and the carrier network. For most London offices, this is a fibre circuit delivered by BT Openreach, Virgin Media Business, Zayo, CityFibre or one of a dozen other carriers, terminating at an optical distribution frame (ODF) in the building’s comms room. From the ODF, internal single-mode fibre cabling routes to the core switching infrastructure, which distributes connectivity across the building.
The access circuit can become a constraint when it is undersized, congested or lacks the availability required by the application. Performance also depends on carrier routing, peering, security services, the data-centre network and the application itself, so the last mile should be measured as part of the complete path.
Where remote services support critical processes, define the acceptable outage, recovery objective and degraded mode before selecting the connectivity design. A second circuit may be justified, but its value depends on genuine diversity and the failover architecture.
A resilient design may use two carrier services, diverse physical routes, separate building entry points and independent network equipment. The required level should follow a business-impact and single-point-of-failure assessment, and route diversity should be verified rather than assumed from different carrier names.
The capacity and resilience of the last-mile connection is the limiting factor for AI application performance in most commercial buildings.
Dark fibre and the direct data centre connection
For some organisations, network latency and consistency to a co-location or cloud location can affect user experience or process performance. The acceptable threshold is application-specific. Measure the current path and agree a target with the application owner before selecting carrier Ethernet, internet, wavelength or dark-fibre services.
Dark fibre is unlit optical fibre leased or provided for the customer to light. It can offer control over capacity and optical equipment, but it does not automatically guarantee a particular latency or resilience level. Route, distance, splices, equipment, operations and service support still need to be designed and managed.
The building-side design normally includes a protected route from the entry point to the optical equipment, suitable single-mode fibre or provider handoff, patching, power, earthing and documented optical-loss limits. The exact design should follow the carrier or optical-system specification.
If the last-mile connection is saturated by routine business traffic, AI application performance degrades regardless of the data centre behind it.
The internal fibre backbone: the overlooked constraint
Most discussions of last-mile connectivity focus on the external circuit. The internal fibre backbone, the cabling between the comms room and telecommunications rooms on each floor, or between multiple comms rooms in a larger building, is often the actual constraint.
An older internal backbone may constrain a new external service if its fibre type, connector condition, distance or installed loss cannot support the intended Ethernet application. Test and identify the existing fibre before selecting transceivers or deciding that replacement is required.
A backbone upgrade can often be phased, but the programme depends on routes, riser access, fire stopping, live services, fibre count and change windows. Confirm the delivery programme only after survey and migration planning.
London and the surrounding region contain several major data-centre clusters. Provider presence and site naming change over time, so shortlist facilities from current carrier, cloud and co-location information rather than embedding an unsupported operator list in the infrastructure specification.
Geographical distance alone does not determine round-trip latency. Carrier route, optical path, switching, security services and application architecture all contribute. Test the actual service or obtain a committed performance specification where latency is material.
Standards and sources
- Uptime Institute Tier StandardData-centre availability and topology context
- IEEE 802.3 EthernetEthernet physical-layer applications and media
- Fluke Networks fibre testingOptical loss, certification and troubleshooting guidance
Frequently asked questions
What is the last-mile connection and why does it matter for AI applications?
The last mile is the access path between the building and the wider carrier network. It can affect throughput, latency and availability, but it is one part of the end-to-end path. Size and test the service against the application, then review carrier routing, resilience, internal backbone capacity and the destination network.
What is dark fibre and is it appropriate for commercial buildings?
Dark fibre is unlit optical fibre that the customer or service provider equips with optical transmission hardware. It can suit organisations that need scalable capacity or greater control, but it brings design, operations and support responsibilities. Compare it with managed wavelengths, carrier Ethernet and internet services using cost, route, capacity, availability and operational capability.
What fibre type is required for a 10G or higher internal building backbone?
The answer depends on the Ethernet application, transceiver, distance and loss budget. OM4 can support specified multimode applications over defined distances, while OS2 is commonly selected for longer reach and future flexibility. Test the installed fibre and use the relevant IEEE and manufacturer limits. Do not apply a universal 300 m rule or assume every OM1/OM2 link is limited to 1G without checking the actual application distance.
How should a commercial building specify resilient last-mile connectivity?
Start with the business impact and recovery objective. Identify route, duct, entry-point, carrier, power and equipment single points of failure. Where justified, use diverse services and independent equipment with tested failover. Do not prescribe a universal one-minute target; the acceptable failover time depends on the applications and continuity design.